This two-year, six-site, competitive renewal compares four strategies for the treatment of SSRI-resistant adolescent depression. We propose to complete our recruitment of 400 adolescents, aged 12-18, who continue to be depressed despite an adequate trial of one of the SSRI antidepressants currently used in community practice (i.e., fluoxetine, citalopram, escitalopram, sertraline). During this project period, we will empanel 100 adolescents with SSRI-resistant depression to complete our target recruitment. The rates of clinically acceptable response (defined as a CGI-I < 2, CGIS < 3, and > 50% decrease in the CDRS-R), slope of decrease of depressive symptoms on the CDRS-R, and direct cost of treatment per unit of improvement will be compared across the 4 cells in a 2x2 factorial design: (1) switch within SSRI class (those on citalopram switch to fluoxetine; those on fluoxetine switch to citalopram; those on sertraline switch randomly to either fluoxetine or citalopram); (2) switch to a different class of agent (venlafaxine); (3) switch within SSRI class plus receive cognitive behavior therapy (CBT); and (4) switch to a different class of agent (venlafaxine) plus CBT. "Adequate treatment" with an SSRI is defined as > 8 weeks of total treatment, at least 4 weeks of which were at the equivalent of 20 mg of fluoxetine, and at least 4 weeks of which were at the equivalent > 40 mg of fluoxetine. In order to enter the study, subjects are assessed at two points in time 2 weeks apart, and must demonstrate sufficient severity (CDRS-R > 40, CGI-S > 4) and stability (decline in CDRS over two weeks < 30%). Subjects meeting these criteria will be cross-tapered gradually from their current regimen. The acute phase of treatment is 12 weeks, during which subjects can receive up to 40 mg of fluoxetine or citalopram, or 225 mg of venlafaxine. Subjects who show a clinically acceptable response will receive continuation treatment for an additional 12 weeks. All subjects, regardless of treatment compliance, will be assessed at intake, midpoint, and the end of acute treatment, and every 6 months thereafter for 1 year after the end of continuation treatment. We hypothesize that, with regard to adequate clinical response and rate of decline of depressive symptoms, venlafaxine will be superior to SSRI switch, CBT + medication will be superior to medication alone, and that CBT + venlafaxine will be superior to the other three cells. Significance. Because 40% of depressed adolescents do not respond to initial treatment with an SSRI, this is a common and serious public health problem. Subjects empaneled thus far are a high-risk group, with prevalent chronic depression and suicidality at intake. This is the only extant study to address the treatment of SSRI-resistant adolescent depression, and will be uniquely informative about treatment guidelines, the relative safety and efficacy of different antidepressants, and the value of adding CBT to medication treatment for this condition. This site is based in Pittsburgh, which also serves as the Data Coordinating Center for the study [unreadable] [unreadable]